Showing 51–100 of 159 results for author: Eisner, J

A Generative Model for Punctuation in Dependency Trees

Authors: Xiang Lisa Li, Dingquan Wang, Jason Eisner

Abstract: Treebanks traditionally treat punctuation marks as ordinary words, but linguists have suggested that a tree's "true" punctuation marks are not observed (Nunberg, 1990). These latent "underlying" marks serve to delimit or separate constituents in the syntax tree. When the tree's yield is rendered as a written sentence, a string rewriting mechanism transduces the underlying marks into "surface" mark… ▽ More Treebanks traditionally treat punctuation marks as ordinary words, but linguists have suggested that a tree's "true" punctuation marks are not observed (Nunberg, 1990). These latent "underlying" marks serve to delimit or separate constituents in the syntax tree. When the tree's yield is rendered as a written sentence, a string rewriting mechanism transduces the underlying marks into "surface" marks, which are part of the observed (surface) string but should not be regarded as part of the tree. We formalize this idea in a generative model of punctuation that admits efficient dynamic programming. We train it without observing the underlying marks, by locally maximizing the incomplete data likelihood (similarly to EM). When we use the trained model to reconstruct the tree's underlying punctuation, the results appear plausible across 5 languages, and in particular, are consistent with Nunberg's analysis of English. We show that our generative model can be used to beat baselines on punctuation restoration. Also, our reconstruction of a sentence's underlying punctuation lets us appropriately render the surface punctuation (via our trained underlying-to-surface mechanism) when we syntactically transform the sentence. △ Less

Submitted 26 June, 2019; originally announced June 2019.

What Kind of Language Is Hard to Language-Model?

Authors: Sabrina J. Mielke, Ryan Cotterell, Kyle Gorman, Brian Roark, Jason Eisner

Abstract: How language-agnostic are current state-of-the-art NLP tools? Are there some types of language that are easier to model with current methods? In prior work (Cotterell et al., 2018) we attempted to address this question for language modeling, and observed that recurrent neural network language models do not perform equally well over all the high-resource European languages found in the Europarl cor… ▽ More How language-agnostic are current state-of-the-art NLP tools? Are there some types of language that are easier to model with current methods? In prior work (Cotterell et al., 2018) we attempted to address this question for language modeling, and observed that recurrent neural network language models do not perform equally well over all the high-resource European languages found in the Europarl corpus. We speculated that inflectional morphology may be the primary culprit for the discrepancy. In this paper, we extend these earlier experiments to cover 69 languages from 13 language families using a multilingual Bible corpus. Methodologically, we introduce a new paired-sample multiplicative mixed-effects model to obtain language difficulty coefficients from at-least-pairwise parallel corpora. In other words, the model is aware of inter-sentence variation and can handle missing data. Exploiting this model, we show that "translationese" is not any easier to model than natively written language in a fair comparison. Trying to answer the question of what features difficult languages have in common, we try and fail to reproduce our earlier (Cotterell et al., 2018) observation about morphological complexity and instead reveal far simpler statistics of the data that seem to drive complexity in a much larger sample. △ Less

Submitted 25 February, 2020; v1 submitted 11 June, 2019; originally announced June 2019.

Comments: Published at ACL 2019

Dusty disk winds at the sublimation rim of the highly inclined, low mass YSO SU Aurigae

Authors: Aaron Labdon, Stefan Kraus, Claire L Davies, Alexander Kreplin, Jacques Kluska, Tim J Harries, John D Monnier, Theo ten Brummelaar, Fabian Baron, Rafael Millan-Gabet, Brian Kloppenborg, Joshua Eisner, Judit Sturmann, Laszlo Sturmann

Abstract: T Tauri stars are low-mass young stars whose disks provide the setting for planet formation. Despite this, their structure is poorly understood. We present new infrared interferometric observations of the SU Aurigae circumstellar environment that offer 3 x higher resolution and better baseline position angle coverage over previous observations. We investigate the characteristics of circumstellar m… ▽ More T Tauri stars are low-mass young stars whose disks provide the setting for planet formation. Despite this, their structure is poorly understood. We present new infrared interferometric observations of the SU Aurigae circumstellar environment that offer 3 x higher resolution and better baseline position angle coverage over previous observations. We investigate the characteristics of circumstellar material around SU Aur, constrain the disk geometry, composition and inner dust rim structure. The CHARA array offers opportunities for long baseline observations, with baselines up to 331 m. Using the CLIMB 3-telescope combiner in the K-band allows us to measure visibilities as well as closure phase. We undertook image reconstruction for model-independent analysis, and geometric modeling. Additionally, the fitting of radiative transfer models constrains the physical parameters of the disk. For the first time, a dusty disk wind is introduced to the radiative transfer code TORUS to model protoplanetary disks. Our implementation is motivated by theoretical dusty disk winds, where magnetic field lines drive dust above the disk plane close to the sublimation zone. Image reconstruction reveals an inclined disk with slight asymmetry along its minor-axis, likely due to inclination effects obscuring the inner disk rim through absorption of incident star light on the near-side and thermal re-emission/scattering of the far-side. Geometric modelling of a skewed ring finds the inner rim at 0.17+/-0.02 au with an inclination of 50.9+/-1.0 degrees and minor axis position angle 60.8+/-1.2 degrees. Radiative transfer modelling shows a flared disk with an inner radius at 0.18 au which implies a grain size of 0.4 um and a scale height of 15.0 au at 100 au. Among the tested radiative transfer models, only the dusty disk wind successfully accounts for the K-band excess by introducing dust above the mid-plane. △ Less

Submitted 28 May, 2019; originally announced May 2019.

Comments: Accepted for publication in Astronomy \& Astrophysics

Journal ref: A&A 627, A36 (2019)

Imputing Missing Events in Continuous-Time Event Streams

Authors: Hongyuan Mei, Guanghui Qin, Jason Eisner

Abstract: Events in the world may be caused by other, unobserved events. We consider sequences of events in continuous time. Given a probability model of complete sequences, we propose particle smoothing---a form of sequential importance sampling---to impute the missing events in an incomplete sequence. We develop a trainable family of proposal distributions based on a type of bidirectional continuous-time… ▽ More Events in the world may be caused by other, unobserved events. We consider sequences of events in continuous time. Given a probability model of complete sequences, we propose particle smoothing---a form of sequential importance sampling---to impute the missing events in an incomplete sequence. We develop a trainable family of proposal distributions based on a type of bidirectional continuous-time LSTM: Bidirectionality lets the proposals condition on future observations, not just on the past as in particle filtering. Our method can sample an ensemble of possible complete sequences (particles), from which we form a single consensus prediction that has low Bayes risk under our chosen loss metric. We experiment in multiple synthetic and real domains, using different missingness mechanisms, and modeling the complete sequences in each domain with a neural Hawkes process (Mei & Eisner 2017). On held-out incomplete sequences, our method is effective at inferring the ground-truth unobserved events, with particle smoothing consistently improving upon particle filtering. △ Less

Submitted 14 May, 2019; originally announced May 2019.

Comments: ICML 2019 camera-ready. The first version of this work appeared on OpenReview in September 2018

Contextualization of Morphological Inflection

Authors: Ekaterina Vylomova, Ryan Cotterell, Timothy Baldwin, Trevor Cohn, Jason Eisner

Abstract: Critical to natural language generation is the production of correctly inflected text. In this paper, we isolate the task of predicting a fully inflected sentence from its partially lemmatized version. Unlike traditional morphological inflection or surface realization, our task input does not provide ``gold'' tags that specify what morphological features to realize on each lemmatized word; rather,… ▽ More Critical to natural language generation is the production of correctly inflected text. In this paper, we isolate the task of predicting a fully inflected sentence from its partially lemmatized version. Unlike traditional morphological inflection or surface realization, our task input does not provide ``gold'' tags that specify what morphological features to realize on each lemmatized word; rather, such features must be inferred from sentential context. We develop a neural hybrid graphical model that explicitly reconstructs morphological features before predicting the inflected forms, and compare this to a system that directly predicts the inflected forms without relying on any morphological annotation. We experiment on several typologically diverse languages from the Universal Dependencies treebanks, showing the utility of incorporating linguistically-motivated latent variables into NLP models. △ Less

Submitted 3 May, 2019; originally announced May 2019.

Comments: NAACL 2019

Protoplanetary Disk Masses from Radiative Transfer Modeling: A Case Study in Taurus

Authors: Nicholas P. Ballering, Josh A. Eisner

Abstract: Measuring the masses of protoplanetary disks is crucial for understanding their planet-forming potential. Typically, dust masses are derived from (sub-)millimeter flux density measurements plus assumptions for the opacity, temperature, and optical depth of the dust. Here we use radiative transfer models to quantify the validity of these assumptions with the aim of improving the accuracy of disk du… ▽ More Measuring the masses of protoplanetary disks is crucial for understanding their planet-forming potential. Typically, dust masses are derived from (sub-)millimeter flux density measurements plus assumptions for the opacity, temperature, and optical depth of the dust. Here we use radiative transfer models to quantify the validity of these assumptions with the aim of improving the accuracy of disk dust mass measurements. We first carry out a controlled exploration of disk parameter space. We find that the disk temperature is a strong function of disk size, while the optical depth depends on both disk size and dust mass. The millimeter-wavelength spectral index can be significantly shallower than the naive expectation due to a combination of optical depth and deviations from the Rayleigh-Jeans regime. We fit radiative transfer models to the spectral energy distributions (SEDs) of 132 disks in the Taurus-Auriga region using a Markov chain Monte Carlo approach. We used all available data to produce the most complete SEDs used in any extant modeling study. We perform the fitting twice: first with unconstrained disk sizes and again imposing the disk size--brightness relation inferred for sources in Taurus. This constraint generally forces the disks to be smaller, warmer, and more optically thick. From both sets of fits, we find disks to be $\sim$1--5 times more massive than when derived using (sub-)millimeter measurements and common assumptions. With the uncertainties derived from our model fitting, the previously measured dust mass--stellar mass correlation is present in our study but only significant at the 2$σ$ level. △ Less

Submitted 19 March, 2019; originally announced March 2019.

Comments: 28 pages, 13 figures, accepted for publication in AJ

High Precision Dynamical Masses of Pre-Main Sequence Stars with ALMA and Gaia

Authors: Patrick D. Sheehan, Ya-Lin Wu, Josh A. Eisner, John Tobin

Abstract: The Keplerian rotation in protoplanetary disks can be used to robustly measure stellar masses at very high precision if the source distance is known. We present Atacama Large Millimeter/submillimeter Array (ALMA) observations of spatially and spectrally resolved $^{12}$CO (2--1) emission towards the disks around 2MASS J16262774--2527247 (the tertiary companion to ROXs 12 at 5100 au), CT Cha, and D… ▽ More The Keplerian rotation in protoplanetary disks can be used to robustly measure stellar masses at very high precision if the source distance is known. We present Atacama Large Millimeter/submillimeter Array (ALMA) observations of spatially and spectrally resolved $^{12}$CO (2--1) emission towards the disks around 2MASS J16262774--2527247 (the tertiary companion to ROXs 12 at 5100 au), CT Cha, and DH Tau. We employ detailed modeling of the Keplerian rotation profile, coupled with accurate distances from $Gaia$, to directly measure the stellar masses with $\sim$2\% precision. We also compare these direct mass measurements with the masses inferred from evolutionary models, determined in a statistically rigorous way. We find that 2MASS J16262774--2527247 has a mass of $0.535^{+0.006}_{-0.007}~M_{\odot}$ and CT Cha has a mass of $0.796^{+0.015}_{-0.014}~M_{\odot}$, broadly consistent with evolutionary models, although potentially significant differences remain. DH Tau has a mass of $0.101^{+0.004}_{-0.003}~M_{\odot}$, but it suffers from strong foreground absorption that may affect our mass estimate. The combination of ALMA, $Gaia$, and codes like \texttt{pdspy}, presented here, can be used to infer the dynamical masses for large samples of young stars and substellar objects, and place constraints on evolutionary models. △ Less

Submitted 28 February, 2019; originally announced March 2019.

Comments: 22 pages, 16 figures, 3 tables, accepted for publication in ApJ

Continuous Schauder frames for Banach spaces

Authors: Joseph Eisner, Daniel Freeman

Abstract: We introduce the notion of a continuous Schauder frame for a Banach space. This is both a generalization of continuous frames and coherent states for Hilbert spaces and a generalization of unconditional Schauder frames for Banach spaces. As a natural example, we prove that any wavelet for $L_p(\R)$ with $1<p<\infty$ generates a continuous wavelet Schauder frame. Furthermore, we generalize the prop… ▽ More We introduce the notion of a continuous Schauder frame for a Banach space. This is both a generalization of continuous frames and coherent states for Hilbert spaces and a generalization of unconditional Schauder frames for Banach spaces. As a natural example, we prove that any wavelet for $L_p(\R)$ with $1<p<\infty$ generates a continuous wavelet Schauder frame. Furthermore, we generalize the properties shrinking and boundedly complete to the continuous Schauder frame setting, and prove that many of the fundamental James theorems still hold in this general context. △ Less

Submitted 20 December, 2018; originally announced December 2018.

Comments: 20 pages

MSC Class: 42C15; 81R30; 46B10

UniMorph 2.0: Universal Morphology

Authors: Christo Kirov, Ryan Cotterell, John Sylak-Glassman, Géraldine Walther, Ekaterina Vylomova, Patrick Xia, Manaal Faruqui, Sabrina J. Mielke, Arya D. McCarthy, Sandra Kübler, David Yarowsky, Jason Eisner, Mans Hulden

Abstract: The Universal Morphology UniMorph project is a collaborative effort to improve how NLP handles complex morphology across the world's languages. The project releases annotated morphological data using a universal tagset, the UniMorph schema. Each inflected form is associated with a lemma, which typically carries its underlying lexical meaning, and a bundle of morphological features from our schema.… ▽ More The Universal Morphology UniMorph project is a collaborative effort to improve how NLP handles complex morphology across the world's languages. The project releases annotated morphological data using a universal tagset, the UniMorph schema. Each inflected form is associated with a lemma, which typically carries its underlying lexical meaning, and a bundle of morphological features from our schema. Additional supporting data and tools are also released on a per-language basis when available. UniMorph is based at the Center for Language and Speech Processing (CLSP) at Johns Hopkins University in Baltimore, Maryland and is sponsored by the DARPA LORELEI program. This paper details advances made to the collection, annotation, and dissemination of project resources since the initial UniMorph release described at LREC 2016. lexical resources} } △ Less

Submitted 25 February, 2020; v1 submitted 25 October, 2018; originally announced October 2018.

Comments: LREC 2018

The LEECH Exoplanet Imaging Survey: Limits on Planet Occurrence Rates Under Conservative Assumptions

Authors: Jordan M. Stone, Andrew J. Skemer, Philip M. Hinz, Mariangela Bonavita, Kaitlin M. Kratter, Anne-Lise Maire, Denis Defrere, Vanessa P. Bailey, Eckhart Spalding, Jarron M. Leisenring, S. Desidera, M. Bonnefoy, Beth Biller, Charles E. Woodward, Th. Henning, Michael F. Skrutskie, J. A. Eisner, Justin R. Crepp, Jennifer Patience, Gerd Weigelt, Robert J. De Rosa, Joshua Schlieder, Wolfgang Brandner, Dániel Apai, Kate Su , et al. (11 additional authors not shown)

Abstract: We present the results of the largest $L^{\prime}$ ($3.8~μ$m) direct imaging survey for exoplanets to date, the Large Binocular Telescope Interferometer (LBTI) Exozodi Exoplanet Common Hunt (LEECH). We observed 98 stars with spectral types from B to M. Cool planets emit a larger share of their flux in $L^{\prime}$ compared to shorter wavelengths, affording LEECH an advantage in detecting low-mass,… ▽ More We present the results of the largest $L^{\prime}$ ($3.8~μ$m) direct imaging survey for exoplanets to date, the Large Binocular Telescope Interferometer (LBTI) Exozodi Exoplanet Common Hunt (LEECH). We observed 98 stars with spectral types from B to M. Cool planets emit a larger share of their flux in $L^{\prime}$ compared to shorter wavelengths, affording LEECH an advantage in detecting low-mass, old, and cold-start giant planets. We emphasize proximity over youth in our target selection, probing physical separations smaller than other direct imaging surveys. For FGK stars, LEECH outperforms many previous studies, placing tighter constraints on the hot-start planet occurrence frequency interior to $\sim20$ au. For less luminous, cold-start planets, LEECH provides the best constraints on giant-planet frequency interior to $\sim20$ au around FGK stars. Direct imaging survey results depend sensitively on both the choice of evolutionary model (e.g., hot- or cold-start) and assumptions (explicit or implicit) about the shape of the underlying planet distribution, in particular its radial extent. Artificially low limits on the planet occurrence frequency can be derived when the shape of the planet distribution is assumed to extend to very large separations, well beyond typical protoplanetary dust-disk radii ($\lesssim50$ au), and when hot-start models are used exclusively. We place a conservative upper limit on the planet occurrence frequency using cold-start models and planetary population distributions that do not extend beyond typical protoplanetary dust-disk radii. We find that $\lesssim90\%$ of FGK systems can host a 7 to 10 $M_{\mathrm{Jup}}$ planet from 5 to 50 au. This limit leaves open the possibility that planets in this range are common. △ Less

Submitted 6 December, 2018; v1 submitted 24 October, 2018; originally announced October 2018.

Comments: 31 pages, 13 figures, accepted to AJ

The CoNLL--SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection

Authors: Ryan Cotterell, Christo Kirov, John Sylak-Glassman, Géraldine Walther, Ekaterina Vylomova, Arya D. McCarthy, Katharina Kann, Sabrina J. Mielke, Garrett Nicolai, Miikka Silfverberg, David Yarowsky, Jason Eisner, Mans Hulden

Abstract: The CoNLL--SIGMORPHON 2018 shared task on supervised learning of morphological generation featured data sets from 103 typologically diverse languages. Apart from extending the number of languages involved in earlier supervised tasks of generating inflected forms, this year the shared task also featured a new second task which asked participants to inflect words in sentential context, similar to a… ▽ More The CoNLL--SIGMORPHON 2018 shared task on supervised learning of morphological generation featured data sets from 103 typologically diverse languages. Apart from extending the number of languages involved in earlier supervised tasks of generating inflected forms, this year the shared task also featured a new second task which asked participants to inflect words in sentential context, similar to a cloze task. This second task featured seven languages. Task 1 received 27 submissions and task 2 received 6 submissions. Both tasks featured a low, medium, and high data condition. Nearly all submissions featured a neural component and built on highly-ranked systems from the earlier 2017 shared task. In the inflection task (task 1), 41 of the 52 languages present in last year's inflection task showed improvement by the best systems in the low-resource setting. The cloze task (task 2) proved to be difficult, and few submissions managed to consistently improve upon both a simple neural baseline system and a lemma-repeating baseline. △ Less

Submitted 25 February, 2020; v1 submitted 16 October, 2018; originally announced October 2018.

Comments: CoNLL 2018. arXiv admin note: text overlap with arXiv:1706.09031

The Wide Integral Field Infrared Spectrograph: Commissioning Results and On-sky Performance

Authors: Suresh Sivanandam, Dae-Sik Moon, R. Elliot Meyer, Jason Grunhut, Dennis Zaritsky, Joshua Eisner, Ke Ma, Charles Henderson, Basil Blank, Chueh-Yi Chou, Miranda E. Jarvis, Stephen Eikenberry, Moo-Young Chun, Byeong-Gon Park

Abstract: We have recently commissioned a novel infrared ($0.9-1.7$ $μ$m) integral field spectrograph (IFS) called the Wide Integral Field Infrared Spectrograph (WIFIS). WIFIS is a unique instrument that offers a very large field-of-view (50$^{\prime\prime}$ x 20$^{\prime\prime}$) on the 2.3-meter Bok telescope at Kitt Peak, USA for seeing-limited observations at moderate spectral resolving power. The measu… ▽ More We have recently commissioned a novel infrared ($0.9-1.7$ $μ$m) integral field spectrograph (IFS) called the Wide Integral Field Infrared Spectrograph (WIFIS). WIFIS is a unique instrument that offers a very large field-of-view (50$^{\prime\prime}$ x 20$^{\prime\prime}$) on the 2.3-meter Bok telescope at Kitt Peak, USA for seeing-limited observations at moderate spectral resolving power. The measured spatial sampling scale is $\sim1\times1^{\prime\prime}$ and its spectral resolving power is $R\sim2,500$ and $3,000$ in the $zJ$ ($0.9-1.35$ $μ$m) and $H_{short}$ ($1.5-1.7$ $μ$m) modes, respectively. WIFIS's corresponding etendue is larger than existing near-infrared (NIR) IFSes, which are mostly designed to work with adaptive optics systems and therefore have very narrow fields. For this reason, this instrument is specifically suited for studying very extended objects in the near-infrared such as supernovae remnants, galactic star forming regions, and nearby galaxies, which are not easily accessible by other NIR IFSes. This enables scientific programs that were not originally possible, such as detailed surveys of a large number of nearby galaxies or a full accounting of nucleosynthetic yields of Milky Way supernova remnants. WIFIS is also designed to be easily adaptable to be used with larger telescopes. In this paper, we report on the overall performance characteristics of the instrument, which were measured during our commissioning runs in the second half of 2017. We present measurements of spectral resolving power, image quality, instrumental background, and overall efficiency and sensitivity of WIFIS and compare them with our design expectations. Finally, we present a few example observations that demonstrate WIFIS's full capability to carry out infrared imaging spectroscopy of extended objects, which is enabled by our custom data reduction pipeline. △ Less

Submitted 10 September, 2018; originally announced September 2018.

Comments: Published in the Proceedings of SPIE Astronomical Telescopes and Instrumentation 2018. 17 pages, 13 figures

On the Complexity and Typology of Inflectional Morphological Systems

Authors: Ryan Cotterell, Christo Kirov, Mans Hulden, Jason Eisner

Abstract: We quantify the linguistic complexity of different languages' morphological systems. We verify that there is an empirical trade-off between paradigm size and irregularity: a language's inflectional paradigms may be either large in size or highly irregular, but never both. Our methodology measures paradigm irregularity as the entropy of the surface realization of a paradigm -- how hard it is to joi… ▽ More We quantify the linguistic complexity of different languages' morphological systems. We verify that there is an empirical trade-off between paradigm size and irregularity: a language's inflectional paradigms may be either large in size or highly irregular, but never both. Our methodology measures paradigm irregularity as the entropy of the surface realization of a paradigm -- how hard it is to jointly predict all the surface forms of a paradigm. We estimate this by a variational approximation. Our measurements are taken on large morphological paradigms from 31 typologically diverse languages. △ Less

Submitted 7 July, 2018; originally announced July 2018.

Comments: TACL 2018

A Deep Generative Model of Vowel Formant Typology

Authors: Ryan Cotterell, Jason Eisner

Abstract: What makes some types of languages more probable than others? For instance, we know that almost all spoken languages contain the vowel phoneme /i/; why should that be? The field of linguistic typology seeks to answer these questions and, thereby, divine the mechanisms that underlie human language. In our work, we tackle the problem of vowel system typology, i.e., we propose a generative probabilit… ▽ More What makes some types of languages more probable than others? For instance, we know that almost all spoken languages contain the vowel phoneme /i/; why should that be? The field of linguistic typology seeks to answer these questions and, thereby, divine the mechanisms that underlie human language. In our work, we tackle the problem of vowel system typology, i.e., we propose a generative probability model of which vowels a language contains. In contrast to previous work, we work directly with the acoustic information -- the first two formant values -- rather than modeling discrete sets of phonemic symbols (IPA). We develop a novel generative probability model and report results based on a corpus of 233 languages. △ Less

Submitted 7 July, 2018; originally announced July 2018.

Comments: NAACL 2018

Are All Languages Equally Hard to Language-Model?

Authors: Ryan Cotterell, Sabrina J. Mielke, Jason Eisner, Brian Roark

Abstract: For general modeling methods applied to diverse languages, a natural question is: how well should we expect our models to work on languages with differing typological profiles? In this work, we develop an evaluation framework for fair cross-linguistic comparison of language models, using translated text so that all models are asked to predict approximately the same information. We then conduct a s… ▽ More For general modeling methods applied to diverse languages, a natural question is: how well should we expect our models to work on languages with differing typological profiles? In this work, we develop an evaluation framework for fair cross-linguistic comparison of language models, using translated text so that all models are asked to predict approximately the same information. We then conduct a study on 21 languages, demonstrating that in some languages, the textual expression of the information is harder to predict with both $n$-gram and LSTM language models. We show complex inflectional morphology to be a cause of performance differences among languages. △ Less

Submitted 25 February, 2020; v1 submitted 10 June, 2018; originally announced June 2018.

Comments: Published at NAACL 2018

Unsupervised Disambiguation of Syncretism in Inflected Lexicons

Authors: Ryan Cotterell, Christo Kirov, Sabrina J. Mielke, Jason Eisner

Abstract: Lexical ambiguity makes it difficult to compute various useful statistics of a corpus. A given word form might represent any of several morphological feature bundles. One can, however, use unsupervised learning (as in EM) to fit a model that probabilistically disambiguates word forms. We present such an approach, which employs a neural network to smoothly model a prior distribution over feature bu… ▽ More Lexical ambiguity makes it difficult to compute various useful statistics of a corpus. A given word form might represent any of several morphological feature bundles. One can, however, use unsupervised learning (as in EM) to fit a model that probabilistically disambiguates word forms. We present such an approach, which employs a neural network to smoothly model a prior distribution over feature bundles (even rare ones). Although this basic model does not consider a token's context, that very property allows it to operate on a simple list of unigram type counts, partitioning each count among different analyses of that unigram. We discuss evaluation metrics for this novel task and report results on 5 languages. △ Less

Submitted 25 February, 2020; v1 submitted 10 June, 2018; originally announced June 2018.

Comments: Published at NAACL 2018

The Eccentric Cavity, Triple Rings, Two-Armed Spirals, and Double Clumps of the MWC 758 Disk

Authors: Ruobing Dong, Sheng-yuan Liu, Josh Eisner, Sean Andrews, Jeffrey Fung, Zhaohuan Zhu, Eugene Chiang, Jun Hashimoto, Hauyu Baobab Liu, Simon Casassus, Thomas Esposito, Yasuhiro Hasegawa, Takayuki Muto, Yaroslav Pavlyuchenkov, David Wilner, Eiji Akiyama, Motohide Tamura, John Wisniewski

Abstract: Spatially resolved structures in protoplanetary disks hint at unseen planets. Previous imaging observations of the transitional disk around MWC 758 revealed an inner cavity, a ring-like outer disk, emission clumps, and spiral arms, all possibly generated by companions. We present ALMA dust continuum observations of MWC 758 at 0.87 millimeter (mm) wavelength with 43$\times$39 mas angular resolution… ▽ More Spatially resolved structures in protoplanetary disks hint at unseen planets. Previous imaging observations of the transitional disk around MWC 758 revealed an inner cavity, a ring-like outer disk, emission clumps, and spiral arms, all possibly generated by companions. We present ALMA dust continuum observations of MWC 758 at 0.87 millimeter (mm) wavelength with 43$\times$39 mas angular resolution (6.9$\times$6.2 AU) and 20 $μ$Jy beam$^{-1}$ rms. The central sub-mm emission cavity is revealed to be eccentric; once deprojected, its outer edge can be well-fitted by an ellipse with an eccentricity of 0.1 and one focus on the star. The broad ring-like outer disk is resolved into three narrow rings with two gaps in between. The outer two rings tentatively show the same eccentricity and orientation as the innermost ring bounding the inner cavity. The two previously known dust emission clumps are resolved in both the radial and azimuthal directions, with radial widths equal to $\sim$4$\times$ the local scale height. Only one of the two spiral arms previously imaged in near-infrared (NIR) scattered light is revealed in ALMA dust emission, at a slightly larger stellocentric distance owing to projection effects. We also submit evidence of disk truncation at $\sim$100 AU based on comparing NIR imaging observations with models. The spirals, the north clump, and the truncated disk edge are all broadly consistent with the presence of one companion exterior to the spirals at roughly 100 AU. △ Less

Submitted 30 May, 2018; originally announced May 2018.

Comments: accepted by ApJ; the fits file of the ALMA image will be made available by ApJ

Protoplanetary Disk Properties in the Orion Nebula Cluster: Initial Results from Deep, High-Resolution ALMA Observations

Authors: J. A. Eisner, H. G. Arce, N. P. Ballering, J. Bally, S. M. Andrews, R. D. Boyden, J. Di Francesco, M. Fang, D. Johnstone, J. S. Kim, R. K. Mann, B. Matthews, I. Pascucci, L. Ricci, P. D. Sheehan, J. P. Williams

Abstract: We present ALMA 850 $μ$m continuum observations of the Orion Nebula Cluster that provide the highest angular resolution ($\sim 0\rlap{.}''1 \approx 40$ AU) and deepest sensitivity ($\sim 0.1$ mJy) of the region to date. We mosaicked a field containing $\sim 225$ optical or near-IR-identified young stars, $\sim 60$ of which are also optically-identified "proplyds". We detect continuum emission at 8… ▽ More We present ALMA 850 $μ$m continuum observations of the Orion Nebula Cluster that provide the highest angular resolution ($\sim 0\rlap{.}''1 \approx 40$ AU) and deepest sensitivity ($\sim 0.1$ mJy) of the region to date. We mosaicked a field containing $\sim 225$ optical or near-IR-identified young stars, $\sim 60$ of which are also optically-identified "proplyds". We detect continuum emission at 850 $μ$m towards $\sim 80$% of the proplyd sample, and $\sim 50$% of the larger sample of previously-identified cluster members. Detected objects have fluxes of $\sim 0.5$-80 mJy. We remove sub-mm flux due to free-free emission in some objects, leaving a sample of sources detected in dust emission. Under standard assumptions of isothermal, optically thin disks, sub-mm fluxes correspond to dust masses of $\sim 0.5$ to 80 Earth masses. We measure the distribution of disk sizes, and find that disks in this region are particularly compact. Such compact disks are likely to be significantly optically thick. The distributions of sub-mm flux and inferred disk size indicate smaller, lower-flux disks than in lower-density star-forming regions of similar age. Measured disk flux is correlated weakly with stellar mass, contrary to studies in other star forming regions that found steeper correlations. We find a correlation between disk flux and distance from the massive star $θ^1$ Ori C, suggesting that disk properties in this region are influenced strongly by the rich cluster environment. △ Less

Submitted 9 May, 2018; originally announced May 2018.

Comments: Accepted for publication in ApJ

Neural Particle Smoothing for Sampling from Conditional Sequence Models

Authors: Chu-Cheng Lin, Jason Eisner

Abstract: We introduce neural particle smoothing, a sequential Monte Carlo method for sampling annotations of an input string from a given probability model. In contrast to conventional particle filtering algorithms, we train a proposal distribution that looks ahead to the end of the input string by means of a right-to-left LSTM. We demonstrate that this innovation can improve the quality of the sample. To… ▽ More We introduce neural particle smoothing, a sequential Monte Carlo method for sampling annotations of an input string from a given probability model. In contrast to conventional particle filtering algorithms, we train a proposal distribution that looks ahead to the end of the input string by means of a right-to-left LSTM. We demonstrate that this innovation can improve the quality of the sample. To motivate our formal choices, we explain how our neural model and neural sampler can be viewed as low-dimensional but nonlinear approximations to working with HMMs over very large state spaces. △ Less

Submitted 28 April, 2018; originally announced April 2018.

Comments: NAACL 2018

On the Diachronic Stability of Irregularity in Inflectional Morphology

Authors: Ryan Cotterell, Christo Kirov, Mans Hulden, Jason Eisner

Abstract: Many languages' inflectional morphological systems are replete with irregulars, i.e., words that do not seem to follow standard inflectional rules. In this work, we quantitatively investigate the conditions under which irregulars can survive in a language over the course of time. Using recurrent neural networks to simulate language learners, we test the diachronic relation between frequency of wor… ▽ More Many languages' inflectional morphological systems are replete with irregulars, i.e., words that do not seem to follow standard inflectional rules. In this work, we quantitatively investigate the conditions under which irregulars can survive in a language over the course of time. Using recurrent neural networks to simulate language learners, we test the diachronic relation between frequency of words and their irregularity. △ Less

Submitted 23 April, 2018; originally announced April 2018.

Comments: accepted to NAACL 2018; withdrawn in order to add more thorough experiments (coming in next version)

Spell Once, Summon Anywhere: A Two-Level Open-Vocabulary Language Model

Authors: Sabrina J. Mielke, Jason Eisner

Abstract: We show how the spellings of known words can help us deal with unknown words in open-vocabulary NLP tasks. The method we propose can be used to extend any closed-vocabulary generative model, but in this paper we specifically consider the case of neural language modeling. Our Bayesian generative story combines a standard RNN language model (generating the word tokens in each sentence) with an RNN-b… ▽ More We show how the spellings of known words can help us deal with unknown words in open-vocabulary NLP tasks. The method we propose can be used to extend any closed-vocabulary generative model, but in this paper we specifically consider the case of neural language modeling. Our Bayesian generative story combines a standard RNN language model (generating the word tokens in each sentence) with an RNN-based spelling model (generating the letters in each word type). These two RNNs respectively capture sentence structure and word structure, and are kept separate as in linguistics. By invoking the second RNN to generate spellings for novel words in context, we obtain an open-vocabulary language model. For known words, embeddings are naturally inferred by combining evidence from type spelling and token context. Comparing to baselines (including a novel strong baseline), we beat previous work and establish state-of-the-art results on multiple datasets. △ Less

Submitted 25 February, 2020; v1 submitted 22 April, 2018; originally announced April 2018.

Comments: Accepted for publication at AAAI 2019

Multiple Gaps in the Disk of the Class I Protostar GY 91

Authors: Patrick Sheehan, Josh Eisner

Abstract: We present the highest spatial resolution ALMA observations to date of the Class I protostar GY 91 in the $ρ$ Ophiuchus L1688 molecular cloud complex. Our 870 $μ$m and 3 mm dust continuum maps show that the GY 91 disk has a radius of $\sim$80 AU, and an inclination of $\sim$40$^{\circ}$, but most interestingly that the disk has three dark lanes located at 10 AU, 40 AU, and 70 AU. We model these fe… ▽ More We present the highest spatial resolution ALMA observations to date of the Class I protostar GY 91 in the $ρ$ Ophiuchus L1688 molecular cloud complex. Our 870 $μ$m and 3 mm dust continuum maps show that the GY 91 disk has a radius of $\sim$80 AU, and an inclination of $\sim$40$^{\circ}$, but most interestingly that the disk has three dark lanes located at 10 AU, 40 AU, and 70 AU. We model these features assuming they are gaps in the disk surface density profile and find that their widths are 7 AU, 30 AU, and 10 AU. These gaps bear a striking resemblance to the gaps seen in the HL Tau disk, suggesting that there may be Saturn-mass planets hiding in the disk. To constrain the relative ages of GY 91 and HL Tau, we also model the disk and envelope of HL Tau and find that they are of similar ages, although GY 91 may be younger. Although snow lines and magnetic dead zones can also produce dark lanes, if planets are indeed carving these gaps then Saturn-mass planets must form within the first $\sim$0.5 Myr of the lifetime of protoplanetary disks. △ Less

Submitted 7 March, 2018; originally announced March 2018.

Comments: 9 pages, 6 figures, 2 tables. Accepted for publication in ApJ

Disk Masses for Embedded Class I Protostars in the Taurus Molecular Cloud

Authors: Patrick D. Sheehan, Josh A. Eisner

Abstract: Class I protostars are thought to represent an early stage in the lifetime of protoplanetary disks, when they are still embedded in their natal envelope. Here we measure the disk masses of 10 Class I protostars in the Taurus Molecular Cloud to constrain the initial mass budget for forming planets in disks. We use radiative transfer modeling to produce synthetic protostar observations and fit the m… ▽ More Class I protostars are thought to represent an early stage in the lifetime of protoplanetary disks, when they are still embedded in their natal envelope. Here we measure the disk masses of 10 Class I protostars in the Taurus Molecular Cloud to constrain the initial mass budget for forming planets in disks. We use radiative transfer modeling to produce synthetic protostar observations and fit the models to a multi-wavelength dataset using a Markov Chain Monte Carlo fitting procedure. We fit these models simultaneously to our new CARMA 1.3 mm observations that are sensitive to the wide range of spatial scales that are expected from protostellar disks and envelopes so as to be able to distinguish each component, as well as broadband spectral energy distributions compiled from the literature. We find a median disk mass of 0.018 M$_{\odot}$ on average, more massive than the Taurus Class II disks, which have median disk mass of $\sim$0.0025 M$_{\odot}$. This decrease in disk mass can be explained if dust grains have grown by a factor of 75 in grain size, indicating that by the Class II stage, at a few Myr, a significant amount of dust grain processing has occurred. However, there is evidence that significant dust processing has occurred even during the Class I stage, so it is likely that the initial mass budget is higher than the value quoted here. △ Less

Submitted 6 December, 2017; originally announced December 2017.

Comments: 29 pages, 5 figures, 2 tables, ApJ, in press

An Explanation of the Very Low Radio Flux of Young Planet-mass Companions

Authors: Ya-Lin Wu, Laird M. Close, Josh A. Eisner, Patrick D. Sheehan

Abstract: We report Atacama Large Millimeter/submillimeter Array (ALMA) 1.3 mm continuum upper limits for 5 planetary-mass companions DH Tau B, CT Cha B, GSC 6214-210 B, 1RXS 1609 B, and GQ Lup B. Our survey, together with other ALMA studies, have yielded null results for disks around young planet-mass companions and placed stringent dust mass upper limits, typically less than 0.1 M_earth, when assuming dus… ▽ More We report Atacama Large Millimeter/submillimeter Array (ALMA) 1.3 mm continuum upper limits for 5 planetary-mass companions DH Tau B, CT Cha B, GSC 6214-210 B, 1RXS 1609 B, and GQ Lup B. Our survey, together with other ALMA studies, have yielded null results for disks around young planet-mass companions and placed stringent dust mass upper limits, typically less than 0.1 M_earth, when assuming dust continuum is optically thin. Such low-mass gas/dust content can lead to a disk lifetime estimate (from accretion rates) much shorter than the age of the system. To alleviate this timescale discrepancy, we suggest that disks around wide companions might be very compact and optically thick, in order to sustain a few Myr of accretion yet have very weak (sub)millimeter flux so as to still be elusive to ALMA. Our order-of-magnitude estimate shows that compact optically-thick disks might be smaller than 1000 R_jup and only emit ~micro-Jy of flux in the (sub)millimeter, but their average temperature can be higher than that of circumstellar disks. The high disk temperature could impede satellite formation, but it also suggests that mid- to far-infrared might be more favorable than radio wavelengths to characterize disk properties. Finally, the compact disk size might imply that dynamical encounters between the companion and the star, or any other scatterers in the system, play a role in the formation of planetary-mass companions. △ Less

Submitted 20 October, 2017; originally announced October 2017.

Comments: Accepted for publication in AJ

Fine-Grained Prediction of Syntactic Typology: Discovering Latent Structure with Supervised Learning

Authors: Dingquan Wang, Jason Eisner

Abstract: We show how to predict the basic word-order facts of a novel language given only a corpus of part-of-speech (POS) sequences. We predict how often direct objects follow their verbs, how often adjectives follow their nouns, and in general the directionalities of all dependency relations. Such typological properties could be helpful in grammar induction. While such a problem is usually regarded as un… ▽ More We show how to predict the basic word-order facts of a novel language given only a corpus of part-of-speech (POS) sequences. We predict how often direct objects follow their verbs, how often adjectives follow their nouns, and in general the directionalities of all dependency relations. Such typological properties could be helpful in grammar induction. While such a problem is usually regarded as unsupervised learning, our innovation is to treat it as supervised learning, using a large collection of realistic synthetic languages as training data. The supervised learner must identify surface features of a language's POS sequence (hand-engineered or neural features) that correlate with the language's deeper structure (latent trees). In the experiment, we show: 1) Given a small set of real languages, it helps to add many synthetic languages to the training data. 2) Our system is robust even when the POS sequences include noise. 3) Our system on this task outperforms a grammar induction baseline by a large margin. △ Less

Submitted 10 October, 2017; originally announced October 2017.

Comments: 16 pages, 5 figures

Journal ref: Transactions of the Association of Computational Linguistics (TACL), 5:147--161, 2017

The Galactic Dependencies Treebanks: Getting More Data by Synthesizing New Languages

Authors: Dingquan Wang, Jason Eisner

Abstract: We release Galactic Dependencies 1.0---a large set of synthetic languages not found on Earth, but annotated in Universal Dependencies format. This new resource aims to provide training and development data for NLP methods that aim to adapt to unfamiliar languages. Each synthetic treebank is produced from a real treebank by stochastically permuting the dependents of nouns and/or verbs to match the… ▽ More We release Galactic Dependencies 1.0---a large set of synthetic languages not found on Earth, but annotated in Universal Dependencies format. This new resource aims to provide training and development data for NLP methods that aim to adapt to unfamiliar languages. Each synthetic treebank is produced from a real treebank by stochastically permuting the dependents of nouns and/or verbs to match the word order of other real languages. We discuss the usefulness, realism, parsability, perplexity, and diversity of the synthetic languages. As a simple demonstration of the use of Galactic Dependencies, we consider single-source transfer, which attempts to parse a real target language using a parser trained on a "nearby" source language. We find that including synthetic source languages somewhat increases the diversity of the source pool, which significantly improves results for most target languages. △ Less

Submitted 10 October, 2017; originally announced October 2017.

Comments: 16 pages, 5 figures

Journal ref: Transactions of the Association of Computational Linguistics (TACL), 4:491--505, 2016

Data Reduction and Image Reconstruction Techniques for Non-Redundant Masking

Authors: Steph Sallum, Josh Eisner

Abstract: The technique of non-redundant masking (NRM) transforms a conventional telescope into an interferometric array. In practice, this provides a much better constrained point spread function than a filled aperture and thus higher resolution than traditional imaging methods. Here we describe an NRM data reduction pipeline. We discuss strategies for NRM observations regarding dithering patterns and cali… ▽ More The technique of non-redundant masking (NRM) transforms a conventional telescope into an interferometric array. In practice, this provides a much better constrained point spread function than a filled aperture and thus higher resolution than traditional imaging methods. Here we describe an NRM data reduction pipeline. We discuss strategies for NRM observations regarding dithering patterns and calibrator selection. We describe relevant image calibrations and use example Large Binocular Telescope datasets to show their effects on the scatter in the Fourier measurements. We also describe the various ways to calculate Fourier quantities, and discuss different calibration strategies. We present the results of image reconstructions from simulated observations where we adjust prior images, weighting schemes, and error bar estimation. We compare two imaging algorithms and discuss implications for reconstructing images from real observations. Finally, we explore how the current state of the art compares to next generation Extremely Large Telescopes. △ Less

Submitted 25 October, 2017; v1 submitted 4 October, 2017; originally announced October 2017.

Comments: 26 pages, 31 figures, accepted for publication in ApJS

CoNLL-SIGMORPHON 2017 Shared Task: Universal Morphological Reinflection in 52 Languages

Authors: Ryan Cotterell, Christo Kirov, John Sylak-Glassman, Géraldine Walther, Ekaterina Vylomova, Patrick Xia, Manaal Faruqui, Sandra Kübler, David Yarowsky, Jason Eisner, Mans Hulden

Abstract: The CoNLL-SIGMORPHON 2017 shared task on supervised morphological generation required systems to be trained and tested in each of 52 typologically diverse languages. In sub-task 1, submitted systems were asked to predict a specific inflected form of a given lemma. In sub-task 2, systems were given a lemma and some of its specific inflected forms, and asked to complete the inflectional paradigm by… ▽ More The CoNLL-SIGMORPHON 2017 shared task on supervised morphological generation required systems to be trained and tested in each of 52 typologically diverse languages. In sub-task 1, submitted systems were asked to predict a specific inflected form of a given lemma. In sub-task 2, systems were given a lemma and some of its specific inflected forms, and asked to complete the inflectional paradigm by predicting all of the remaining inflected forms. Both sub-tasks included high, medium, and low-resource conditions. Sub-task 1 received 24 system submissions, while sub-task 2 received 3 system submissions. Following the success of neural sequence-to-sequence models in the SIGMORPHON 2016 shared task, all but one of the submissions included a neural component. The results show that high performance can be achieved with small training datasets, so long as models have appropriate inductive bias or make use of additional unlabeled data or synthetic data. However, different biasing and data augmentation resulted in disjoint sets of inflected forms being predicted correctly, suggesting that there is room for future improvement. △ Less

Submitted 4 July, 2017; v1 submitted 27 June, 2017; originally announced June 2017.

Comments: CoNLL 2017

Improved Constraints on the Disk Around MWC 349A from the 23-Meter LBTI

Authors: Steph Sallum, Josh Eisner, Phil Hinz, Patrick Sheehan, Andy Skemer, Peter Tuthill, John Young

Abstract: We present new spatially resolved observations of MWC 349A from the Large Binocular Telescope Interferometer (LBTI), a 23-meter baseline interferometer made up of two, co-mounted 8-meter telescopes. MWC 349A is a B[e] star with an unknown evolutionary state. Proposed scenarios range from a young stellar object (YSO), to a B[e] supergiant, to a tight binary system. Radio continuum and recombination… ▽ More We present new spatially resolved observations of MWC 349A from the Large Binocular Telescope Interferometer (LBTI), a 23-meter baseline interferometer made up of two, co-mounted 8-meter telescopes. MWC 349A is a B[e] star with an unknown evolutionary state. Proposed scenarios range from a young stellar object (YSO), to a B[e] supergiant, to a tight binary system. Radio continuum and recombination line observations of this source revealed a sub-arcsecond bipolar outflow surrounding a ~100 mas circumstellar disk. Followup infrared studies detected the disk, and suggested that it may have skew and an inner clearing. Our new infrared interferometric observations, which have more than twice the resolution of previously-published datasets, support the presence of both skew and a compact infrared excess. They rule out inner clearings with radii greater than ~14 mas. We show the improvements in disk parameter constraints provided by LBTI, and discuss the inferred disk parameters in the context of the posited evolutionary states for MWC 349A. △ Less

Submitted 15 June, 2017; originally announced June 2017.

Comments: 11 pages, 7 figures, accepted for publication in ApJ

Probabilistic Typology: Deep Generative Models of Vowel Inventories

Authors: Ryan Cotterell, Jason Eisner

Abstract: Linguistic typology studies the range of structures present in human language. The main goal of the field is to discover which sets of possible phenomena are universal, and which are merely frequent. For example, all languages have vowels, while most---but not all---languages have an /u/ sound. In this paper we present the first probabilistic treatment of a basic question in phonological typology:… ▽ More Linguistic typology studies the range of structures present in human language. The main goal of the field is to discover which sets of possible phenomena are universal, and which are merely frequent. For example, all languages have vowels, while most---but not all---languages have an /u/ sound. In this paper we present the first probabilistic treatment of a basic question in phonological typology: What makes a natural vowel inventory? We introduce a series of deep stochastic point processes, and contrast them with previous computational, simulation-based approaches. We provide a comprehensive suite of experiments on over 200 distinct languages. △ Less

Submitted 3 May, 2017; originally announced May 2017.

Comments: ACL 2017

WL 17: A Young Embedded Transition Disk

Authors: Patrick D. Sheehan, Josh A. Eisner

Abstract: We present the highest spatial resolution ALMA observations to date of the Class I protostar WL 17 in the $ρ$ Ophiuchus L1688 molecular cloud complex, which show that it has a 12 AU hole in the center of its disk. We consider whether WL 17 is actually a Class II disk being extincted by foreground material, but find that such models do not provide a good fit to the broadband SED and also require su… ▽ More We present the highest spatial resolution ALMA observations to date of the Class I protostar WL 17 in the $ρ$ Ophiuchus L1688 molecular cloud complex, which show that it has a 12 AU hole in the center of its disk. We consider whether WL 17 is actually a Class II disk being extincted by foreground material, but find that such models do not provide a good fit to the broadband SED and also require such high extinction that it would presumably arise from dense material close to the source such as a remnant envelope. Self-consistent models of a disk embedded in a rotating collapsing envelope can nicely reproduce both the ALMA 3 mm observations and the broadband SED of WL 17. This suggests that WL 17 is a disk in the early stages of its formation, and yet even at this young age the inner disk has been depleted. Although there are multiple pathways for such a hole to be created in a disk, if this hole were produced by the formation of planets it could place constraints on the timescale for the growth of planetesimals in protoplanetary disks. △ Less

Submitted 25 April, 2017; originally announced April 2017.

Comments: 7 pages, 3 figures, 2 tables, accepted for publication in ApJL

The ALMA View of the OMC1 Explosion in Orion

Authors: John Bally, Adam Ginsburg, Hector Arce, Josh Eisner, Allison Youngblood, Luis Zapata, Hans Zinnecker

Abstract: Most massive stars form in dense clusters where gravitational interactions with other stars may be common. The two nearest forming massive stars, the BN object and Source I, located behind the Orion Nebula, were ejected with velocities of $\sim$29 and $\sim$13 km s$^{-1}$ about 500 years ago by such interactions. This event generated an explosion in the gas. New ALMA observations show in unprecede… ▽ More Most massive stars form in dense clusters where gravitational interactions with other stars may be common. The two nearest forming massive stars, the BN object and Source I, located behind the Orion Nebula, were ejected with velocities of $\sim$29 and $\sim$13 km s$^{-1}$ about 500 years ago by such interactions. This event generated an explosion in the gas. New ALMA observations show in unprecedented detail, a roughly spherically symmetric distribution of over a hundred $^{12}$CO J=2$-$1 streamers with velocities extending from V$_{LSR}$ =$-$150 to +145 km s$^{-1}$. The streamer radial velocities increase (or decrease) linearly with projected distance from the explosion center, forming a `Hubble Flow' confined to within 50 arcseconds of the explosion center. They point toward the high proper-motion, shock-excited H$_2$ and [Fe ii ] `fingertips' and lower-velocity CO in the H$_2$ wakes comprising Orion's `fingers'. In some directions, the H$_2$ `fingers' extend more than a factor of two farther from the ejection center than the CO streamers. Such deviations from spherical symmetry may be caused by ejecta running into dense gas or the dynamics of the N-body interaction that ejected the stars and produced the explosion. This $\sim$10$^{48}$ erg event may have been powered by the release of gravitational potential energy associated with the formation of a compact binary or a protostellar merger. Orion may be the prototype for a new class of stellar explosion responsible for luminous infrared transients in nearby galaxies. △ Less

Submitted 1 February, 2017; v1 submitted 7 January, 2017; originally announced January 2017.

Comments: 34 pages, 17 figures

The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process

Authors: Hongyuan Mei, Jason Eisner

Abstract: Many events occur in the world. Some event types are stochastically excited or inhibited---in the sense of having their probabilities elevated or decreased---by patterns in the sequence of previous events. Discovering such patterns can help us predict which type of event will happen next and when. We model streams of discrete events in continuous time, by constructing a neurally self-modulating mu… ▽ More Many events occur in the world. Some event types are stochastically excited or inhibited---in the sense of having their probabilities elevated or decreased---by patterns in the sequence of previous events. Discovering such patterns can help us predict which type of event will happen next and when. We model streams of discrete events in continuous time, by constructing a neurally self-modulating multivariate point process in which the intensities of multiple event types evolve according to a novel continuous-time LSTM. This generative model allows past events to influence the future in complex and realistic ways, by conditioning future event intensities on the hidden state of a recurrent neural network that has consumed the stream of past events. Our model has desirable qualitative properties. It achieves competitive likelihood and predictive accuracy on real and synthetic datasets, including under missing-data conditions. △ Less

Submitted 21 November, 2017; v1 submitted 29 December, 2016; originally announced December 2016.

Comments: NIPS 2017 camera-ready. New experiments including intensity prediction evaluation, sensitivity to # of parameters, training speed analysis. Results updated to use final test data instead of devtest. Improved exposition, especially of continuous-time LSTM and thinning algorithm

iLocater: A Diffraction-limited Doppler Spectrometer for the Large Binocular Telescope

Authors: Justin R. Crepp, Jonathan Crass, David King, Andrew Bechter, Eric Bechter, Ryan Ketterer, Robert Reynolds, Philip Hinz, Derek Kopon, David Cavalieri, Louis Fantano, Corina Koca, Eleanya Onuma, Karl Stapelfeldt, Joseph Thomes, Sheila Wall, Steven Macenka, James McGuire, Ronald Korniski, Leonard Zugby, Joshua Eisner, B. Scott Gaudi, Fred Hearty, Kaitlin Kratter, Marc Kuchner , et al. (9 additional authors not shown)

Abstract: We are develo** a stable and precise spectrograph for the Large Binocular Telescope (LBT) named "iLocater." The instrument comprises three principal components: a cross-dispersed echelle spectrograph that operates in the YJ-bands (0.97-1.30 microns), a fiber-injection acquisition camera system, and a wavelength calibration unit. iLocater will deliver high spectral resolution (R~150,000-240,000)… ▽ More We are develo** a stable and precise spectrograph for the Large Binocular Telescope (LBT) named "iLocater." The instrument comprises three principal components: a cross-dispersed echelle spectrograph that operates in the YJ-bands (0.97-1.30 microns), a fiber-injection acquisition camera system, and a wavelength calibration unit. iLocater will deliver high spectral resolution (R~150,000-240,000) measurements that permit novel studies of stellar and substellar objects in the solar neighborhood including extrasolar planets. Unlike previous planet-finding instruments, which are seeing-limited, iLocater operates at the diffraction limit and uses single mode fibers to eliminate the effects of modal noise entirely. By receiving starlight from two 8.4m diameter telescopes that each use "extreme" adaptive optics (AO), iLocater shows promise to overcome the limitations that prevent existing instruments from generating sub-meter-per-second radial velocity (RV) precision. Although optimized for the characterization of low-mass planets using the Doppler technique, iLocater will also advance areas of research that involve crowded fields, line-blanketing, and weak absorption lines. △ Less

Submitted 14 September, 2016; originally announced September 2016.

Comments: 13 pages, 11 figures

Journal ref: Proc. SPIE 9908, Ground-based and Airborne Instrumentation for Astronomy VI, 990819 (4 August 2016)

Imaging protoplanets: observing transition disks with non-redundant masking

Authors: Steph Sallum, Josh Eisner, Laird M. Close, Philip M. Hinz, Katherine B. Follette, Kaitlin Kratter, Andrew J. Skemer, Vanessa P. Bailey, Runa Briguglio, Denis Defrere, Bruce A. Macintosh, Jared R. Males, Katie M. Morzinski, Alfio T. Puglisi, Timothy J. Rodigas, Eckhart Spalding, Peter G. Tuthill, Amali Vaz, Alycia Weinberger, Marco Xomperio

Abstract: Transition disks, protoplanetary disks with inner clearings, are promising objects in which to directly image forming planets. The high contrast imaging technique of non-redundant masking is well posed to detect planetary mass companions at several to tens of AU in nearby transition disks. We present non-redundant masking observations of the T Cha and LkCa 15 transition disks, both of which host p… ▽ More Transition disks, protoplanetary disks with inner clearings, are promising objects in which to directly image forming planets. The high contrast imaging technique of non-redundant masking is well posed to detect planetary mass companions at several to tens of AU in nearby transition disks. We present non-redundant masking observations of the T Cha and LkCa 15 transition disks, both of which host posited sub-stellar mass companions. However, due to a loss of information intrinsic to the technique, observations of extended sources (e.g. scattered light from disks) can be misinterpreted as moving companions. We discuss tests to distinguish between these two scenarios, with applications to the T Cha and LkCa 15 observations. We argue that a static, forward-scattering disk can explain the T Cha data, while LkCa 15 is best explained by multiple orbiting companions. △ Less

Submitted 8 September, 2016; originally announced September 2016.

Comments: SPIE conference proceeding

Journal ref: Proc. SPIE 9907, Optical and Infrared Interferometry and Imaging V, 99070D (August 4, 2016)

A VLA Survey For Faint Compact Radio Sources in the Orion Nebula Cluster

Authors: Patrick D. Sheehan, Josh A. Eisner, Rita K. Mann, Jonathan P. Williams

Abstract: We present Karl G. Janksy Very Large Array (VLA) 1.3 cm, 3.6 cm, and 6 cm continuum maps of compact radio sources in the Orion Nebular Cluster. We mosaicked 34 square arcminutes at 1.3 cm, 70 square arcminutes at 3.6 cm and 109 square arcminutes at 6 cm, containing 778 near-infrared detected YSOs and 190 HST-identified proplyds (with significant overlap between those characterizations). We detecte… ▽ More We present Karl G. Janksy Very Large Array (VLA) 1.3 cm, 3.6 cm, and 6 cm continuum maps of compact radio sources in the Orion Nebular Cluster. We mosaicked 34 square arcminutes at 1.3 cm, 70 square arcminutes at 3.6 cm and 109 square arcminutes at 6 cm, containing 778 near-infrared detected YSOs and 190 HST-identified proplyds (with significant overlap between those characterizations). We detected radio emission from 175 compact radio sources in the ONC, including 26 sources that were detected for the first time at these wavelengths. For each detected source we fit a simple free-free and dust emission model to characterize the radio emission. We extrapolate the free-free emission spectrum model for each source to ALMA bands to illustrate how these measurements could be used to correctly measure protoplanetary disk dust masses from sub-millimeter flux measurements. Finally, we compare the fluxes measured in this survey with previously measured fluxes for our targets, as well as four separate epochs of 1.3 cm data, to search for and quantify variability of our sources. △ Less

Submitted 2 September, 2016; v1 submitted 31 August, 2016; originally announced September 2016.

Comments: 13 pages, 6 figures, 4 tables, ApJ, in press

Prompt planetesimal formation beyond the snow line

Authors: Philip J. Armitage, Josh A. Eisner, Jacob B. Simon

Abstract: We develop a simple model to predict the radial distribution of planetesimal formation. The model is based on the observed growth of dust to mm-sized particles, which drift radially, pile-up, and form planetesimals where the stop** time and dust-to-gas ratio intersect the allowed region for streaming instability-induced gravitational collapse. Using an approximate analytic treatment, we first sh… ▽ More We develop a simple model to predict the radial distribution of planetesimal formation. The model is based on the observed growth of dust to mm-sized particles, which drift radially, pile-up, and form planetesimals where the stop** time and dust-to-gas ratio intersect the allowed region for streaming instability-induced gravitational collapse. Using an approximate analytic treatment, we first show that drifting particles define a track in metallicity--stop** time space whose only substantial dependence is on the disk's angular momentum transport efficiency. Prompt planetesimal formation is feasible for high particle accretion rates (relative to the gas, $\dot{M}_p / \dot{M} > 3 \times 10^{-2}$ for $α= 10^{-2}$), that could only be sustained for a limited period of time. If it is possible, it would lead to the deposition of a broad and massive belt of planetesimals with a sharp outer edge. Including turbulent diffusion and vapor condensation processes numerically, we find that a modest enhancement of solids near the snow line occurs for cm-sized particles, but that this is largely immaterial for planetesimal formation. We note that radial drift couples planetesimal formation across radii in the disk, and suggest that considerations of planetesimal formation favor a model in which the initial deposition of material for giant planet cores occurs well beyond the snow line. △ Less

Submitted 11 August, 2016; originally announced August 2016.

Comments: ApJ Letters, in press

L-band Spectroscopy with Magellan-AO/Clio2: First Results on Young Low-Mass Companions

Authors: Jordan M. Stone, Josh Eisner, Andy Skemer, Katie M. Morzinski, Laird Close, Jared Males, Timothy J. Rodigas, Phil Hinz, Alfio Puglisi

Abstract: L-band spectroscopy is a powerful probe of cool low-gravity atmospheres: The P, Q, and R branch fundamental transitions of methane near 3.3 $μ$m provide a sensitive probe of carbon chemistry; cloud thickness modifies the spectral slope across the band; and H$_{3}^{+}$ opacity can be used to detect aurorae. Many directly imaged gas-giant companions to nearby young stars exhibit L-band fluxes distin… ▽ More L-band spectroscopy is a powerful probe of cool low-gravity atmospheres: The P, Q, and R branch fundamental transitions of methane near 3.3 $μ$m provide a sensitive probe of carbon chemistry; cloud thickness modifies the spectral slope across the band; and H$_{3}^{+}$ opacity can be used to detect aurorae. Many directly imaged gas-giant companions to nearby young stars exhibit L-band fluxes distinct from the field population of brown dwarfs at the same effective temperature. Here we describe commissioning the L-band spectroscopic mode of Clio2, the 1-5 $μ$m instrument behind the Magellan adaptive-optics system. We use this system to measure L-band spectra of directly imaged companions. Our spectra are generally consistent with the parameters derived from previous near-infrared spectra for these late M to early L type objects. Therefore, deviations from the field sequence are constrained to occur below 1500 K. This range includes the L-T transition for field objects and suggests that observed discrepancies are due to differences in cloud structure and CO/CH$_{4}$ chemistry. △ Less

Submitted 8 August, 2016; originally announced August 2016.

Comments: Accepted to ApJ

Protoplanetary Disks in the Orion OMC1 Region Imaged with ALMA

Authors: J. A. Eisner, J. M. Bally, A. Ginsburg, P. D. Sheehan

Abstract: We present ALMA observations of the Orion Nebula that cover the OMC1 outflow region. Our focus in this paper is on compact emission from protoplanetary disks. We mosaicked a field containing $\sim 600$ near-IR-identified young stars, around which we can search for sub-mm emission tracing dusty disks. Approximately 100 sources are known proplyds identified with HST. We detect continuum emission at… ▽ More We present ALMA observations of the Orion Nebula that cover the OMC1 outflow region. Our focus in this paper is on compact emission from protoplanetary disks. We mosaicked a field containing $\sim 600$ near-IR-identified young stars, around which we can search for sub-mm emission tracing dusty disks. Approximately 100 sources are known proplyds identified with HST. We detect continuum emission at 1 mm wavelengths towards $\sim 20\%$ of the proplyd sample, and $\sim 8\%$ of the larger sample of near-IR objects. The noise in our maps allows 4$σ$ detection of objects brighter than $\sim 1.5$ mJy, corresponding to protoplanetary disk masses larger than 1.5 M$_{\rm J}$ (using standard assumptions about dust opacities and gas-to-dust ratios). None of these disks are detected in contemporaneous CO(2-1) or C$^{18}$O(2-1) observations, suggesting that the gas-to-dust ratios may be substantially smaller than the canonical value of 100. Furthermore, since dust grains may already be sequestered in large bodies in ONC disks, the inferred masses of disk solids may be underestimated. Our results suggest that the distribution of disk masses in this region is compatible with the detection rate of massive planets around M dwarfs, which are the dominant stellar constituent in the ONC. △ Less

Submitted 11 April, 2016; originally announced April 2016.

Comments: Accepted for publication in ApJ

Adaptive Optics imaging of VHS 1256-1257: A Low Mass Companion to a Brown Dwarf Binary System

Authors: Jordan M. Stone, Andrew J. Skemer, Kaitlin M. Kratter, Trent J. Dupuy, Laird M. Close, Josh A. Eisner, Jonathan J. Fortney, Philip M. Hinz, Jared R. Males, Caroline V. Morley, Katie M. Morzinski, Kimberly Ward-Duong

Abstract: Recently, Gauza et al. (2015) reported the discovery of a companion to the late M-dwarf, VHS J125601.92-125723.9 (VHS 1256-1257). The companion's absolute photometry suggests its mass and atmosphere are similar to the HR 8799 planets. However, as a wide companion to a late-type star, it is more accessible to spectroscopic characterization. We discovered that the primary of this system is an equal-… ▽ More Recently, Gauza et al. (2015) reported the discovery of a companion to the late M-dwarf, VHS J125601.92-125723.9 (VHS 1256-1257). The companion's absolute photometry suggests its mass and atmosphere are similar to the HR 8799 planets. However, as a wide companion to a late-type star, it is more accessible to spectroscopic characterization. We discovered that the primary of this system is an equal-magnitude binary. For an age $\sim300$ Myr the A and B components each have a mass of $64.6^{+0.8}_{-2.0}~M_{\mathrm{Jup}}$, and the b component has a mass of $11.2^{+9.7}_{-1.8}$, making VHS 1256-1257 only the third brown dwarf triple system. There exists some tension between the spectrophotometric distance of $17.2\pm2.6$ pc and the parallax distance of $12.7\pm1.0$ pc. At 12.7 pc VHS1256-1257 A and B would be the faintest known M7.5 objects, and are even faint outliers among M8 types. If the larger spectrophotmetric distance is more accurate than the parallax, then the mass of each component increases. In particular, the mass of the b component increases well above the deuterium burning limit to $\sim35~M_{\mathrm{Jup}}$ and the mass of each binary component increases to $73^{+20}_{-17}~M_{\mathrm{Jup}}$. At 17.1 pc, the UVW kinematics of the system are consistent with membership in the AB~Dor moving group. The architecture of the system resembles a hierarchical stellar multiple suggesting it formed via an extension of the star-formation process to low masses. Continued astrometric monitoring will resolve this distance uncertainty and will provide dynamical masses for a new benchmark system. △ Less

Submitted 13 January, 2016; originally announced January 2016.

Comments: Accepted to ApJL

The LEECH Exoplanet Imaging Survey: Characterization of the Coldest Directly Imaged Exoplanet, GJ 504 b, and Evidence for Super-Stellar Metallicity

Authors: Andrew J. Skemer, Caroline V. Morley, Neil T. Zimmerman, Michael F. Skrutskie, Jarron Leisenring, Esther Buenzli, Mickael Bonnefoy, Vanessa Bailey, Philip Hinz, Denis Defrére, Simone Esposito, Dániel Apai, Beth Biller, Wolfgang Brandner, Laird Close, Justin R. Crepp, Robert J. De Rosa, Silvano Desidera, Josh Eisner, Jonathan Fortney, Richard Freedman, Thomas Henning, Karl-Heinz Hofmann, Taisiya Kopytova, Roxana Lupu , et al. (17 additional authors not shown)

Abstract: As gas giant planets and brown dwarfs radiate away the residual heat from their formation, they cool through a spectral type transition from L to T, which encompasses the dissipation of cloud opacity and the appearance of strong methane absorption. While there are hundreds of known T-type brown dwarfs, the first generation of directly-imaged exoplanets were all L-type. Recently, Kuzuhara et al. (2… ▽ More As gas giant planets and brown dwarfs radiate away the residual heat from their formation, they cool through a spectral type transition from L to T, which encompasses the dissipation of cloud opacity and the appearance of strong methane absorption. While there are hundreds of known T-type brown dwarfs, the first generation of directly-imaged exoplanets were all L-type. Recently, Kuzuhara et al. (2013) announced the discovery of GJ 504 b, the first T dwarf exoplanet. GJ 504 b provides a unique opportunity to study the atmosphere of a new type of exoplanet with a ~500 K temperature that bridges the gap between the first directly imaged planets (~1000 K) and our own Solar System's Jupiter (~130 K). We observed GJ 504 b in three narrow L-band filters (3.71, 3.88, and 4.00 microns), spanning the red end of the broad methane fundamental absorption feature (3.3 microns) as part of the LEECH exoplanet imaging survey. By comparing our new photometry and literature photometry to a grid of custom model atmospheres, we were able to fit GJ 504 b's unusual spectral energy distribution for the first time. We find that GJ 504 b is well-fit by models with the following parameters: T_eff=544+/-10 K, g<600 m/s^2, [M/H]=0.60+/-0.12, cloud opacity parameter of f_sed=2-5, R=0.96+/-0.07 R_Jup, and log(L)=-6.13+/-0.03 L_Sun, implying a hot start mass of 3-30 M_jup for a conservative age range of 0.1-6.5 Gyr. Of particular interest, our model fits suggest that GJ 504 b has a super-stellar metallicity. Since planet formation can create objects with non-stellar metallicities, while binary star formation cannot, this result suggests that GJ 504 b formed like a planet, not like a binary companion. △ Less

Submitted 30 November, 2015; originally announced November 2015.

Comments: resubmitted to ApJ with referee's comments

Accreting Protoplanets in the LkCa 15 Transition Disk

Authors: S. Sallum, K. B. Follette, J. A. Eisner, L. M. Close, P. Hinz, K. Kratter, J. Males, A. Skemer, B. Macintosh, P. Tuthill, V. Bailey, D. Defrère, K. Morzinski, T. Rodigas, E. Spalding, A. Vaz, A. J. Weinberger

Abstract: Exoplanet detections have revolutionized astronomy, offering new insights into solar system architecture and planet demographics. While nearly 1900 exoplanets have now been discovered and confirmed, none are still in the process of formation. Transition discs, protoplanetary disks with inner clearings best explained by the influence of accreting planets, are natural laboratories for the study of p… ▽ More Exoplanet detections have revolutionized astronomy, offering new insights into solar system architecture and planet demographics. While nearly 1900 exoplanets have now been discovered and confirmed, none are still in the process of formation. Transition discs, protoplanetary disks with inner clearings best explained by the influence of accreting planets, are natural laboratories for the study of planet formation. Some transition discs show evidence for the presence of young planets in the form of disc asymmetries or infrared sources detected within their clearings, as in the case of LkCa 15. Attempts to observe directly signatures of accretion onto protoplanets have hitherto proven unsuccessful. Here we report adaptive optics observations of LkCa 15 that probe within the disc clearing. With accurate source positions over multiple epochs spanning 2009 - 2015, we infer the presence of multiple companions on Keplerian orbits. We directly detect Hα emission from the innermost companion, LkCa 15 b, evincing hot (~10,000 K) gas falling deep into the potential well of an accreting protoplanet. △ Less

Submitted 23 November, 2015; originally announced November 2015.

Comments: 35 pages, 3 figures, 1 table, 9 extended data items

Journal ref: 2015 Nature 527, 342

The LEECH Exoplanet Imaging Survey: Orbit and Component Masses of the Intermediate Age, Late-Type Binary NO UMa

Authors: Joshua E. Schlieder, Andrew J. Skemer, Anne-Lise Maire, Silvano Desidera, Philip Hinz, Michael F. Skrutskie, Jarron Leisenring, Vanessa Bailey, Denis Defrere, Simone Esposito, Klaus G. Strassmeier, Michael Weber, Beth A. Biller, Mickael Bonnefoy, Esther Buenzli, Laird M. Close, Justin R. Crepp, Josh A. Eisner, Karl-Heinz Hofmann, Thomas Henning, Katie M. Morzinski, Dieter Schertl, Gerd Weigelt, Charles E. Woodward

Abstract: We present high-resolution Large Binocular Telescope LBTI/LMIRcam images of the spectroscopic and astrometric binary NO UMa obtained as part of the LBTI Exozodi Exoplanet Common Hunt (LEECH) exoplanet imaging survey. Our H, K$_s$, and L'-band observations resolve the system at angular separations <0.09". The components exhibit significant orbital motion over a span of ~7 months. We combine our ima… ▽ More We present high-resolution Large Binocular Telescope LBTI/LMIRcam images of the spectroscopic and astrometric binary NO UMa obtained as part of the LBTI Exozodi Exoplanet Common Hunt (LEECH) exoplanet imaging survey. Our H, K$_s$, and L'-band observations resolve the system at angular separations <0.09". The components exhibit significant orbital motion over a span of ~7 months. We combine our imaging data with archival images, published speckle interferometry measurements, and existing spectroscopic velocity data to solve the full orbital solution and estimate component masses. The masses of the K2.0$\pm$0.5 primary and K6.5$\pm$0.5 secondary are 0.83$\pm$0.02 M$_{\odot}$ and 0.64$\pm$0.02 M$_{\odot}$, respectively. We also derive a system distance of d = 25.87$\pm$0.02 pc and revise the Galactic kinematics of NO UMa. Our revised Galactic kinematics confirm NO UMa as a nuclear member of the ~500 Myr old Ursa Major moving group and it is thus a mass and age benchmark. We compare the masses of the NO UMa binary components to those predicted by five sets of stellar evolution models at the age of the Ursa Major group. We find excellent agreement between our measured masses and model predictions with little systematic scatter between the models. NO UMa joins the short list of nearby, bright, late-type binaries having known ages and fully characterized orbits. △ Less

Submitted 13 October, 2015; originally announced October 2015.

Comments: 11 pages, 5 figures, 4 tables, accepted for publication in ApJ

The Early ALMA View of the FU Ori Outburst System

Authors: A. S. Hales, S. A. Corder, W. R. D. Dent, S. M. Andrews, J. A. Eisner, L. A. Cieza

Abstract: We have obtained ALMA Band 7 observations of the FU Ori outburst system at 0.6"x0.5" resolution to measure the link between the inner disk instability and the outer disk through sub-mm continuum and molecular line observations. Our observations detect continuum emission which can be well modeled by two unresolved sources located at the position of each binary component. The interferometric observa… ▽ More We have obtained ALMA Band 7 observations of the FU Ori outburst system at 0.6"x0.5" resolution to measure the link between the inner disk instability and the outer disk through sub-mm continuum and molecular line observations. Our observations detect continuum emission which can be well modeled by two unresolved sources located at the position of each binary component. The interferometric observations recover the entire flux reported in previous single-dish studies, ruling out the presence of a large envelope. Assuming that the dust is optically thin, we derive disk dust masses of $2\times 10^{-4}$M$_{\odot}$ and $8\times 10^{-5}$M$_{\odot}$, for the north and south components respectively. We place limits on the disks' radii of $r<$45 AU. We report the detection of molecular emission from $^{12}$CO(3-2), HCO$^{+}$(4-3) and from HCN(4-3). The $^{12}$CO appears widespread across the two binary components, and is slightly more extended than the continuum emission. The denser gas tracer HCO$^{+}$ peaks close to the position of the southern binary component, while HCN appears peaked at the position of the northern component. This suggests that the southern binary component is embedded in denser molecular material, consistent with previous studies that indicate a heavily reddened object. At this angular resolution any interaction between the two unresolved disk components cannot be disentangled. Higher resolution images are vital to understanding the process of star formation via rapid accretion FU Ori-type episodes. △ Less

Submitted 8 September, 2015; originally announced September 2015.

Comments: ApJ Accepted

Exoplanet science with the LBTI: instrument status and plans

Authors: D. Defrère, P. Hinz, A. Skemer, V. Bailey, E. Downey, O. Durney, J. Eisner, J. M. Hill, W. F. Hoffmann, J. Leisenring, T. McMahon, M. Montoya, E. Spalding, J. Stone, A. Vaz, O. Absil, S. Esposito, M. Kenworthy, B. Mennesson, R. Millan-Gabet, M. Nelson, A. Puglisi, M. F. Skrutskie, J. Wilson

Abstract: The Large Binocular Telescope Interferometer (LBTI) is a strategic instrument of the LBT designed for high-sensitivity, high-contrast, and high-resolution infrared (1.5-13 $μ$m) imaging of nearby planetary systems. To carry out a wide range of high-spatial resolution observations, it can combine the two AO-corrected 8.4-m apertures of the LBT in various ways including direct (non-interferometric)… ▽ More The Large Binocular Telescope Interferometer (LBTI) is a strategic instrument of the LBT designed for high-sensitivity, high-contrast, and high-resolution infrared (1.5-13 $μ$m) imaging of nearby planetary systems. To carry out a wide range of high-spatial resolution observations, it can combine the two AO-corrected 8.4-m apertures of the LBT in various ways including direct (non-interferometric) imaging, coronagraphy (APP and AGPM), Fizeau imaging, non-redundant aperture masking, and nulling interferometry. It also has broadband, narrowband, and spectrally dispersed capabilities. In this paper, we review the performance of these modes in terms of exoplanet science capabilities and describe recent instrumental milestones such as first-light Fizeau images (with the angular resolution of an equivalent 22.8-m telescope) and deep interferometric nulling observations. △ Less

Submitted 3 September, 2015; originally announced September 2015.

Comments: 12 pages, 6 figures, Proc. SPIE

Approximation-Aware Dependency Parsing by Belief Propagation

Authors: Matthew R. Gormley, Mark Dredze, Jason Eisner

Abstract: We show how to train the fast dependency parser of Smith and Eisner (2008) for improved accuracy. This parser can consider higher-order interactions among edges while retaining O(n^3) runtime. It outputs the parse with maximum expected recall -- but for speed, this expectation is taken under a posterior distribution that is constructed only approximately, using loopy belief propagation through str… ▽ More We show how to train the fast dependency parser of Smith and Eisner (2008) for improved accuracy. This parser can consider higher-order interactions among edges while retaining O(n^3) runtime. It outputs the parse with maximum expected recall -- but for speed, this expectation is taken under a posterior distribution that is constructed only approximately, using loopy belief propagation through structured factors. We show how to adjust the model parameters to compensate for the errors introduced by this approximation, by following the gradient of the actual loss on training data. We find this gradient by back-propagation. That is, we treat the entire parser (approximations and all) as a differentiable circuit, as Stoyanov et al. (2011) and Domke (2010) did for loopy CRFs. The resulting trained parser obtains higher accuracy with fewer iterations of belief propagation than one trained by conditional log-likelihood. △ Less

Submitted 10 August, 2015; originally announced August 2015.

ALMA Observations of the Largest Proto-Planetary Disk in the Orion Nebula, 114-426: A CO Silhouette

Authors: John Bally, Rita K. Mann, Josh Eisner, Sean M. Andrews, James Di Francesco, Meredith Hughes, Doug Johnstone, Brenda Matthews, Luca Ricci, Jonathan P. Williams

Abstract: We present ALMA observations of the largest protoplanetary disk in the Orion Nebula, 114-426. Detectable 345 GHz (856 micron) dust continuum is produced only in the 350 AU central region of the ~1000 AU diameter silhouette seen against the bright H-alpha background in HST images. Assuming optically thin dust emission at 345 GHz, a gas-to-dust ratio of 100, and a grain temperature of 20 K, the disk… ▽ More We present ALMA observations of the largest protoplanetary disk in the Orion Nebula, 114-426. Detectable 345 GHz (856 micron) dust continuum is produced only in the 350 AU central region of the ~1000 AU diameter silhouette seen against the bright H-alpha background in HST images. Assuming optically thin dust emission at 345 GHz, a gas-to-dust ratio of 100, and a grain temperature of 20 K, the disk gas-mass is estimated to be 3.1 +/- 0.6 Jupiter masses. If most solids and ices have have been incorporated into large grains, however, this value is a lower limit. The disk is not detected in dense-gas tracers such as HCO+ J=4-3, HCN J=4-3, or CS =7-6. These results may indicate that the 114-426 disk is evolved and depleted in some light organic compounds found in molecular clouds. The CO J=3-2 line is seen in absorption against the bright 50 to 80 K background of the Orion A molecular cloud over the full spatial extent and a little beyond the dust continuum emission. The CO absorption reaches a depth of 27 K below the background CO emission at VLSR ~6.7 km/s about 0.52 arcseconds (210 AU) northeast and 12 K below the background CO emission at VLSR ~ 9.7 km/s about 0.34 arcseconds (140 AU) southwest of the suspected location of the central star, implying that the embedded star has a mass less than 1 Solar mass . △ Less

Submitted 10 June, 2015; originally announced June 2015.

Comments: 20 pages, 4 figures

Spectral Energy Distributions of Accreting Protoplanets

Authors: J. A. Eisner

Abstract: Planets are often invoked as the cause of inferred gaps or inner clearings in transition disks. These putative planets would interact with the remnant circumstellar disk, accreting gas and generating substantial luminosity. Here I explore the expected appearance of accreting protoplanets at a range of evolutionary states. I compare synthetic spectral energy distributions with the handful of claime… ▽ More Planets are often invoked as the cause of inferred gaps or inner clearings in transition disks. These putative planets would interact with the remnant circumstellar disk, accreting gas and generating substantial luminosity. Here I explore the expected appearance of accreting protoplanets at a range of evolutionary states. I compare synthetic spectral energy distributions with the handful of claimed detections of substellar-mass companions in transition disks. While observed fluxes of candidate companions are generally compatible with accreting protoplanets, challenges remain in reconciling the extended structure inferred in observed objects with the compact emission expected from protoplanets or circumplanetary disks. I argue that a large fraction of transition disks should harbor bright protoplanets, and that more may be detected as larger telescopes open up additional parameter space. △ Less

Submitted 18 February, 2015; originally announced February 2015.

Comments: Accepted for publication in ApJ Letters

Protoplanetary Disk Masses in the Young NGC 2024 Cluster

Authors: Rita K. Mann, Sean M. Andrews, Josh A. Eisner, Jonathan P. Williams, Michael R. Meyer, James Di Francesco, John M. Carpenter, Doug Johnstone

Abstract: We present the results from a Submillimeter Array survey of the 887 micron continuum emission from the protoplanetary disks around 95 young stars in the young cluster NGC 2024. Emission was detected from 22 infrared sources, with flux densities from ~5 to 330 mJy; upper limits (at 3sigma) for the other 73 sources range from 3 to 24 mJy. For standard assumptions, the corresponding disk masses range… ▽ More We present the results from a Submillimeter Array survey of the 887 micron continuum emission from the protoplanetary disks around 95 young stars in the young cluster NGC 2024. Emission was detected from 22 infrared sources, with flux densities from ~5 to 330 mJy; upper limits (at 3sigma) for the other 73 sources range from 3 to 24 mJy. For standard assumptions, the corresponding disk masses range from ~0.003 to 0.2Msolar, with upper limits at 0.002--0.01Msolar. The NGC 2024 sample has a slightly more populated tail at the high end of its disk mass distribution compared to other clusters, but without more information on the nature of the sample hosts it remains unclear if this difference is statistically significant or a superficial selection effect. Unlike in the Orion Trapezium, there is no evidence for a disk mass dependence on the (projected) separation from the massive star IRS2b in the NGC 2024 cluster. We suggest that this is due to either the cluster youth or a comparatively weaker photoionizing radiation field. △ Less

Submitted 26 January, 2015; originally announced January 2015.

Comments: ApJ, in press

New Spatially Resolved Observations of the T Cha Transition Disk and Constraints on the Previously Claimed Substellar Companion

Authors: S. Sallum, J. A. Eisner, Laird M. Close, Philip M. Hinz, Andrew J. Skemer, Vanessa Bailey, Runa Briguglio, Katherine B. Follette, Jared R. Males, Katie M. Morzinski, Alfio Puglisi, Timothy J. Rodigas, Alycia J. Weinberger, Marco Xompero

Abstract: We present multi-epoch non-redundant masking observations of the T Cha transition disk, taken at the VLT and Magellan in H, Ks, and L' bands. T Cha is one of a small number of transition disks that host companion candidates discovered by high-resolution imaging techniques, with a putative companion at a position angle of 78 degrees, separation of 62 mas, and contrast at L' of 5.1 mag. We find comp… ▽ More We present multi-epoch non-redundant masking observations of the T Cha transition disk, taken at the VLT and Magellan in H, Ks, and L' bands. T Cha is one of a small number of transition disks that host companion candidates discovered by high-resolution imaging techniques, with a putative companion at a position angle of 78 degrees, separation of 62 mas, and contrast at L' of 5.1 mag. We find comparable binary parameters in our re-reduction of the initial detection images, and similar parameters in the 2011 L', 2013 NaCo L', and 2013 NaCo Ks data sets. We find a close-in companion signal in the 2012 NaCo L' dataset that cannot be explained by orbital motion, and a non-detection in the 2013 MagAO/Clio2 L' data. However, Monte-carlo simulations show that the best fits to the 2012 NaCo and 2013 MagAO/Clio2 followup data may be consistent with noise. There is also a significant probability of false non-detections in both of these data sets. We discuss physical scenarios that could cause the best fits, and argue that previous companion and scattering explanations are inconsistent with the results of the much larger dataset presented here. △ Less

Submitted 8 January, 2015; originally announced January 2015.

Comments: 25 pages, 22 figures, accepted for publication in ApJ